Five Turning Points in the Historical Progress of Statistics - My Personal Vision

Statistics has penetrated almost all branches of science and all areas of human endeavor. At the same time, statistics is not only misunderstood, misused and abused to a frightening extent, but it is also often much disliked by students in colleges and universities. This lecture discusses/covers/addresses the historical development of statistics, aiming at identifying the most important turning points that led to the present state of statistics and at answering the questions “What went wrong with statistics?” and “What to do next?”. ACM Computing Classification System (1998): A.0, A.m, G.3, K.3.2.

Estimating cell diffusivity and cell proliferation rate by interpreting IncuCyte ZOOM™ assay data using the Fisher-Kolmogorov model

Background: Standard methods for quantifying IncuCyte ZOOM™ assays involve measurements that quantify how rapidly the initially-vacant area becomes re-colonised with cells as a function of time. Unfortunately, these measurements give no insight into the details of the cellular-level mechanisms acting to close the initially-vacant area. We provide an alternative method enabling us to quantify the role of cell motility and cell proliferation separately. To achieve this we calibrate standard data available from IncuCyte ZOOM™ images to the solution of the Fisher-Kolmogorov model. Results: The Fisher-Kolmogorov model is a reaction-diffusion equation that has been used to describe collective cell spreading driven by cell migration, characterised by a cell diffusivity, D, and carrying capacity limited proliferation with proliferation rate, λ, and carrying capacity density, K. By analysing temporal changes in cell density in several subregions located well-behind the initial position of the leading edge we estimate λ and K. Given these estimates, we then apply automatic leading edge detection algorithms to the images produced by the IncuCyte ZOOM™ assay and match this data with a numerical solution of the Fisher-Kolmogorov equation to provide an estimate of D. We demonstrate this method by applying it to interpret a suite of IncuCyte ZOOM™ assays using PC-3 prostate cancer cells and obtain estimates of D, λ and K. Comparing estimates of D, λ and K for a control assay with estimates of D, λ and K for assays where epidermal growth factor (EGF) is applied in varying concentrations confirms that EGF enhances the rate of scratch closure and that this stimulation is driven by an increase in D and λ, whereas K is relatively unaffected by EGF. Conclusions: Our approach for estimating D, λ and K from an IncuCyte ZOOM™ assay provides more detail about cellular-level behaviour than standard methods for analysing these assays. In particular, our approach can be used to quantify the balance of cell migration and cell proliferation and, as we demonstrate, allow us to quantify how the addition of growth factors affects these processes individually.

A fully discrete C-0 interior penalty Galerkin approximation of the extended Fisher-Kolmogorov equation

A fully discrete C-0 interior penalty finite element method is proposed and analyzed for the Extended Fisher-Kolmogorov (EFK) equation u(t) + gamma Delta(2)u - Delta u + u(3) - u = 0 with appropriate initial and boundary conditions, where gamma is a positive constant. We derive a regularity estimate for the solution u of the EFK equation that is explicit in gamma and as a consequence we derive a priori error estimates that are robust in gamma. (C) 2013 Elsevier B.V. All rights reserved.

Generalizations and extensions of the Fokker-Planck-Kolmogorov equations

The problem of determining probability density functions of general transformations of random processes is considered in this thesis. A method of solution is developed in which partial differential equations satisfied by the unknown density function are derived. These partial differential equations are interpreted as generalized forms of the classical Fokker-Planck-Kolmogorov equations and are shown to imply the classical equations for certain classes of Markov processes. Extensions of the generalized equations which overcome degeneracy occurring in the steady-state case are also obtained.

The equations of Darling and Siegert are derived as special cases of the generalized equations thereby providing unity to two previously existing theories. A technique for treating non-Markov processes by studying closely related Markov processes is proposed and is seen to yield the Darling and Siegert equations directly from the classical Fokker-Planck-Kolmogorov equations.

As illustrations of their applicability, the generalized Fokker-Planck-Kolmogorov equations are presented for certain joint probability density functions associated with the linear filter. These equations are solved for the density of the output of an arbitrary linear filter excited by Markov Gaussian noise and for the density of the output of an RC filter excited by the Poisson square wave. This latter density is also found by using the extensions of the generalized equations mentioned above. Finally, some new approaches for finding the output probability density function of an RC filter-limiter-RC filter system driven by white Gaussian noise are included. The results in this case exhibit the data required for complete solution and clearly illustrate some of the mathematical difficulties inherent to the use of the generalized equations.

Biological controls on resuspension and the relationship between particle size and the Kolmogorov length scale in a shallow coastal sea

Measurements of suspended particle matter (SPM) and turbulence have been obtained over five tidal surveys during spring and summer 2010 at station L4 (5025 degrees N 04.22 degrees W, depth 50 m), in the Western English Channel. The relationship between turbulence intensity and bed stress is explored, with an in-line holographic imaging system evaluating the extent to which material is resuspended. Image analysis allows for the identification of SPM above a size threshold of 200 pm, capturing particle variability across tidal cycles and the two seasons. Dissipation of turbulent kinetic energy, which exceeds 10(-5) W kg(-1), yields maximum values of bed stress of between 0.17 and 0.20 N m(-2), frequently resulting in the resuspension of material from the bed. Resuspension is shown to promote aggregation of SPM into flocs, where the size of such particles is theoretically determined by the Kolmogorov microscale, l(k). During the spring surveys, flocs of a size larger than lk were observed, though this was not repeated during summer. It is proposed that the presence of gelatinous, biological material in spring allows flocculated particles to exceed l(k). This suggests that under specific circumstances, the limiting factor on the growth of flocculated SPM is not only turbulence, as previously thought, but the presence or absence of certain types of biological particle.

Statistic Complexity: Combining Kolmogorov Complexity with an Ensemble Approach

Background: The evaluation of the complexity of an observed object is an old but outstanding problem. In this paper we are tying on this problem introducing a measure called statistic complexity.

Avaliação da compressão de dados e da qualidade de imagem em modelos de animação gráfica para web : uma nova abordagem baseada em complexidade de Kolmogorov

Este trabalho versa sobre a avaliação da compressão de dados e da qualidade de imagens e animações usando-se complexidade de Kolmogorov, simulação de máquinas e distância de informação. Complexidade de Kolmogorov é uma teoria da informação e da aleatoriedade baseada na máquina de Turing. No trabalho é proposto um método para avaliar a compressão de dados de modelos de animação gráfica usando-se simulação de máquinas. Também definimos formalmente compressão de dados com perdas e propomos a aplicação da distância de informação como uma métrica de qualidade de imagem. O desenvolvimento de uma metodologia para avaliar a compressão de dados de modelos de animação gráfica para web é útil, a medida que as páginas na web estão sendo cada vez mais enriquecidas com animações, som e vídeo, e a economia de banda de canal tornase importante, pois os arquivos envolvidos são geralmente grandes. Boa parte do apelo e das vantagens da web em aplicações como, por exemplo, educação à distância ou publicidade, reside exatamente na existência de elementos multimídia, que apoiam a idéia que está sendo apresentada na página. Como estudo de caso, o método de comparação e avaliação de modelos de animação gráfica foi aplicado na comparação de dois modelos: GIF (Graphics Interchange Format) e AGA (Animação Gráfica baseada em Autômatos finitos), provando formalmente que AGA é melhor que GIF (“melhor” significa que AGA comprime mais as animações que GIF). Foi desenvolvida também uma definição formal de compressão de dados com perdas com o objetivo de estender a metodologia de avalição apresentada Distância de informação é proposta como uma nova métrica de qualidade de imagem, e tem como grande vantagem ser uma medida universal, ou seja, capaz de incorporar toda e qualquer medida computável concebível. A métrica proposta foi testada em uma série de experimentos e comparada com a distância euclidiana (medida tradicionalmente usada nestes casos). Os resultados dos testes são uma evidência prática que a distância proposta é efetiva neste novo contexto de aplicação, e que apresenta, em alguns casos, resultados superiores ao da distância euclidiana. Isto também é uma evidência que a distância de informação é uma métrica mais fina que a distância euclidiana. Também mostramos que há casos em que podemos aplicar a distância de informação, mas não podemos aplicar a distância euclidiana. A métrica proposta foi aplicada também na avaliação de animações gráficas baseadas em frames, onde apresentou resultados melhores que os obtidos com imagens puras. Este tipo de avaliação de animações é inédita na literatura, segundo revisão bibliográfica feita. Finalmente, neste trabalho é apresentado um refinamento à medida proposta que apresentou resultados melhores que a aplicação simples e direta da distância de informação.

Global dynamics of stationary solutions of the extended Fisher-Kolmogorov equation

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Patch-size and isolation effects in the Fisher-Kolmogorov equation

We examine the classical problem of the existence of a threshold size for a patch to allow for survival of a given population in the case where the patch is not completely isolated. The surrounding habitat matrix is characterized by a non-zero carrying capacity. We show that a critical patch size cannot be strictly defined in this case. We also obtain the saturation density in such a patch as a function of the size of the patch and the relative carrying capacity of the outer region. We argue that this relative carrying capacity is a measure of the isolation of the patch. Our results are then compared with conclusions drawn from observations of the population dynamics of understorey birds in fragments of the Amazonian forest and shown to qualitatively agree with them, offering an explanation for the importance of dispersal and isolation in these observations. Finally, we show that a generalized critical patch size can be introduced resorting to threshold densities for the observation of a given species.

The carreau-yasuda fluids: a skin friction equation for turbulent flow in pipes and kolmogorov dissipative scales

In this work the turbulent flow of the Non-Newtonian Carreau-Yasuda fluid will be studied. A skin friction equation for the turbulent flow of Carreau-Yasuda fluids will be derived assuming a logarithmic behavior of the turbulent mean velocity for the near wall flow out of the viscous sub layer. An alternative near wall characteristic length scale which takes into account the effects of the relaxation time will be introduced. The characteristic length will be obtained through the analysis of viscous region near the wall. The results compared with experimental data obtained with Tylose (methyl hydroxil cellulose) solutions showing good agreement. The relations between scales integral and dissipative obtained for length, time, velocity, kinetic energy, and vorticity will be derived for this type of fluid. When the power law index approach to unity the relations reduces to Newtonian case.

Modelo de Fisher-Kolmogorov em dinâmica populacional com capacidade de suporte espacialmente dependente

Pós-graduação em Física - IFT